I am hypothesizing a cross-level interaction for a study and have a question about the most appropriate way to specify the model using HLM Software.
I have an interaction term in my level-1 equation, preceded by the two IVs that make up the interaction term. I am hypothesizing that a level-2 variable will moderate the slope between the interaction term and my DV.
My Question are:
Even though I'm not predicting that the level-2 variable will influence anything other than the slope between the interaction term and my DV, is it proper model specification to include the level-2 variable in all the level-2 equations (predicting the intercept, the slope of IV1 and the DV, and the slope of the IV2 and DV)?
Can I simply include the level-2 variable in only the level-2 equation for the interaction and DV slope?
Also, I do have a significant portion of variance in the slopes for all of the IV-DV relationships.Any suggestions would be greatly appreciated!
Dear trevor,
I think that you have to consult Rabe and Skrondal (Second or Third Version) Multilevel and Longitudinal Modeling Using Stata to model your problem. After this i can help you in modelling this specification.
Good luck
扫码加好友，拉您进群
全部版块
我的主页
收藏
